While you had your wounds mended and gilded with gold over many years through contemplation and healing...
When you were finally able to recover and live a normal life among the masses...
When you had finally forgotten everything from the past...
I showed up...
Leave Me Be (LMB) is an artwork that talks about pain and was created after gathering multiple stories of trauma from multiple people. LMB showcases a human figure generatively created and composed of an ethereal starry field. This figure has a golden line that is randomly generated, in a kintsukuroi fashion, representative of the healed wounds of the person. LMB has a machine-learning based face-detection algorithm which operates through the device camera. Once a face is detected on the camera, the golden healing line changes to a deep red wound from which a trickle of blood flows down, ie., when you stand in front of the art, it starts to bleed.
Final art : https://nurecas.github.io/leavemebe/
(Although the NFT contains the correct working link to the art, due to problems with Pinata and Opensea, the art will mot render correctly in Opensea. Please use the link above to experience the art.)
This artwork was not auctioned in the typical highest bid for money. Instead I used "stories as a currency". I asked the community to share their stories of healing and one among them was chosen by the end of the auction period and gifted with the NFT of the artwork. Details:
Once the art is opened in a web browser, it asks for permission to use the camera. Please note that your camera feed is completely private, ie. it is used only in the browser and is not fed to any server. The code is completely front-end.
It usually takes 10+ seconds to detect the face for the first time. Make sure the face in front of the camera is well lit. Once the camera detects the face, the line colour will change to red and start dripping. If the face is hidden or the camera is covered, the line will change back to the original golden colour.
Following is a demo video showcasing the same:
This work was created using the p5js and ml5js libraries.